Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

One not-so-fun place this could go is mandatory voice recognition commands, leading to everything said in the car being recorded and stored by the manufacturers.


Voice recognition might be the only UI worse for safety and usability than a touchscreen for normal driving operations.

Not that you're wrong about the privacy angle either.


Why? Talking won't prevent you focusing on the road?


Yes: attempting to have conversation is found to diminish focus on driving to a large extent — I remember seeing a study on this, and can vouch with personal experience.

Yes, you can do most of the driving, but "at the edges", when quicker reaction time is needed, it becomes noticeable. Similar to, ahem, drunk driving, though obviously, not as bad, and you can stop a conversation whenever needed.

Obviously, talking to a computer in your car would be less taxing than to a person, but when it misrecognizes the input, it might be the opposite.


Because voice recognition is horribly imprecise. If you're controlling essential functions for driving then you need controls that are efficient, predictable, and reliable.


Sounds like a implementation problem, not a problem with voice control.

We have a 20 old navi with voice control. You can't just say free form things, but it's very deterministic. Most commands you want to say aren't free form, so this doesn't really matter. It also confirms everything, so it will never do something without you knowing. It also has the best voice I got to know. Natural, precise, short AND friendly; no clue why all these modern voices with way more compute all sound like garbage.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: